Hydrodesulfurization of 4,6-dimethyldibenzothiophene over noble metals supported on mesoporous zeolites

Article abstract of DOI:10.1002/anie.200802540

(Figure Presented) Pores for thought: Noble-metal catalysts supported on mesoporous zeolites have been found to be much more efficient than on microporous zeolites and γ-Al₂O₃ for the hydrodesulfurization of 4,6-dimethyldibenzothioph

Full text of DOI:10.1002/anie.200802540

Communications
DOI: 10.1002/anie.200802540
Heterogeneous Catalysis
Hydrodesulfurization of 4,6-Dimethyldibenzothiophene over Noble
Metals Supported on Mesoporous Zeolites**
Yinyong Sun and Roel Prins*
The reduction of the sulfur content in gasoline and diesel fuel
has been a subject of intense investigation in recent years
because the sulfur level must in many countries be reduced to
It is well known that zeolites possess strong acidity, high
stability, and a regular pore array, and are for these reasons
applied in many industrial catalytic reactions. However, their
small pore size means that relatively large molecules such as
4,6-DM-DBT cannot enter the pores; they can only react on
the outer surface of the zeolites and cannot reach many active
centers. A support with strong acidity and relatively large
pores would, therefore, be preferred. The recent discovery of
mesoporous zeolites with their hierarchical porosity and
strong acidity opens the possibility of using them as supports
1
0 ppm by the year 2010 for environmental reasons, while for
fuel-cell applications the sulfur content should be below
[
1]
0
.1 ppm. To reach this low level, even highly refractory
molecules such as 4,6-dimethyldibenzothiophene (4,6-DM-
DBT) must be desulfurized. However, because of steric
hindrance by the methyl groups adjacent to the sulfur atom,
desulfurization of 4,6-DM-DBT mainly takes place after the
[
2]
[9]
molecule has first been hydrogenated. Therefore, the
hydrogenating ability of the catalyst is of critical importance
for deep hydrodesulfurization (HDS). Recent studies have
in HDS. However, until now their use as a support in HDS
has not been reported.
Herein we report on Pt, Pd, and Pt-Pd catalysts supported
on mesoporous Na-ZSM-5. The catalytic activity and selec-
tivity of these catalysts were studied in the HDS of 4,6-DM-
DBT, and the hydrocarbon products as well as the hydro-
genated intermediates were analyzed. Compared with con-
ventional Na-ZSM-5- or g-Al O -supported catalysts, the
[
3]
shown that noble-metal-supported catalysts have much
better hydrogenation performance than conventional metal
sulfides in HDS, and may be used in the second reactor of a
deep HDS process.
Not only the active catalyst, but also the support plays an
important role in the catalytic performance of catalysts.
Acidic supports can increase the conversion of dibenzothio-
2
3
mesoporous Na-ZSM-5-supported catalysts exhibited much
better catalytic performance for hydrodesulfurization.
The powder XRD patterns of mesoporous Na-ZSM-5
(MNZ-5) and Na-ZSM-5 (NZ-5, see Figure S1 in the Sup-
porting Information) show well-resolved peaks which are
characteristic of the ZSM-5 zeolite structure. MNZ-5 exhib-
[
4]
phene (DBT) and of 4,6-DM-DBT. One explanation for this
is that they enable dealkylation and isomerization reactions of
the alkyl substituents, which may transform refractory
components into more reactive species and thus accelerate
[
4b,d]
HDS.
Moreover, acidic supports may also improve the
ited a type IV N adsorption/desorption isotherm (see Figure
2
catalytic activity of the catalyst particles. Since partial
electron transfer can occur from the metal particles to
acidic sites of the support, the resulting electron-deficient
S2a in the Supporting Information) typical for mesoporous
materials. In contrast, NZ-5 showed a type I isotherm, which
is typical of microporous materials. Moreover, a uniform pore
distribution centered at around 4.9 nm was observed for
MNZ-5 (see Figure S2b in the Supporting Information). The
detailed sorption data of both samples are listed in Table 1.
The BET special surface area and mesoporous volume for
[5]
metal particles are deemed to have a better resistance to
[
5b,6]
sulfur poisoning by decreasing the interaction with H S.
2
Another explanation for this improvement is the creation of a
second hydrogenation pathway by spillover of hydrogen
atoms from the metal particles to the aromatic sulfur-
containing molecules that are adsorbed on acidic sites in the
2
À1
3
À1
MNZ-5 are 579 m g and 0.44 cm g , respectively, much
higher than those of NZ-5.
[
7]
vicinity of the metal particles. While the metal particles
become poisoned by sulfur, they can still dissociate hydrogen
molecules, and thus the hydrogenation pathway involving
The isomerization of 2-methyl-2-pentene (2M2P) is a
[10]
good model reaction to evaluate the acidity of solid acids.
The molar ratio of trans-3-methyl-2-pentene (trans-3M2P,
obtained by shift of a methyl group) to trans- and cis-4-
methyl-2-pentene (trans- and cis-4M2P, respectively, obtained
by shift of an H atom) in the product reflects the acidity of
solid acids. The higher the molar ratio is, the stronger the
acidity is. Table 1 shows that MNZ-5 resulted in a higher
conversion in the isomerization of 2M2P than did NZ-5; this
observation may be attributed to the high BET surface area of
MNZ-5. The analysis of the Na content in the two zeolites
showed that only 90% of the Al atoms were charge-compen-
sated by Na cations, so that about 10% protons existed in NZ-
5 and MNZ-5, thus suggesting that they are indeed acidic
[8]
spillover would still be possible.
[
*] Dr. Y. Sun, Prof. Dr. R. Prins
Institute for Chemical and Bioengineering, ETH Zꢀrich
093 Zꢀrich (Switzerland)
8
Fax: (+41)44-632-1162
E-mail: prins@chem.ethz.ch
[
**] We thank Prof. Jeroen A. van Bokhoven and Dr. Huamin Wang for
helpful discussions, and Nadiya Danilina for the characterization of
samples.
supports. g-Al O , on the other hand, gave a lower conversion
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.200802540.
2
3
8478
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 8478 –8481

Angewandte
Chemie
Table 1: Textural parameters and acidity evaluation of various supports.
containing intermediates 4,6-dime-
thyltetrahydrodibenzothiophene
(4,6-TH-DM-DBT) and 4,6-dime-
thylhexahydrodibenzothiophene
Supports
BET area
Mesopore
volume [cm g
Mesopore
diameter [nm]
Conversion
Molar
ratio
2
À1
3
À1
[a]
[b]
[
m g
]
]
[%]
g-Al O3
220
375
579
0.60
0.09
0.44
5.7
–
4.9
21
25
40
0
0.20
0.33
2
(
4,6-HH-DM-DBT), and the desul-
furized products 3,3’-DM-CHB and
,3’-DM-BCH (see Figure S3 in the
NZ-5
MNZ-5
3
[a] Conversion of 2M2P at 423 K. [b] Molar ratio of trans-3M2P to trans- and cis-4M2P.
Supporting Information). The prod-
uct yields and selectivities showed
that 3,3’-DM-CHB was always the
than NZ-5 and MNZ-5, and no trans-3M2P was detected, thus
indicating that g-Al O has a relatively weak acidity.
Figure 1 shows the catalytic activity of Pt catalysts on
various supports in the HDS of 4,6-DM-DBT. The HDS
most abundant product, irrespective of the weight time (see
Figure S4 in the Supporting Information). The yield and
selectivity reached 28 and 41%, respectively, at high weight
time. 3,3’-DM-BCH was the second major product, and had a
similar selectivity as TMC. The yield and selectivity of 3,3’-
DM-BCH were 17 and 25%, respectively, and those of TMC
were 16 and 24%, respectively. The selectivity of 3,3’-DM-BP
2
3
(3%) was constant throughout the reaction, thus indicating
that 3,3’-DM-BP does not hydrogenate further on Pt/MNZ-5
under the HDS conditions. The selectivity of the sulfur-
containing molecules 4,6-DM-TH-DBT and 4,6-DM-HH-
DBT decreased with weight time. They were observed in
small amounts at high weight time, which suggests that these
intermediates can be desulfurized easily. The same sulfur-
containing intermediates and hydrocarbons were observed
with Pt/NZ-5 as with Pt/MNZ-5, but with only a trace amount
of TMC. 3,3’-DM-CHB was also the most abundant product
and showed increasing selectivity with weight time. The yield
and selectivity were 20 and 54%, respectively, at high weight
time. 3,3’-DM-BCH was the second major product, with a
yield and selectivity of 7.7 and 20%, respectively. The
selectivity of 3,3’-DM-BP (5%) was also constant throughout
the reaction and the yield was only 4%. The selectivity of 4,6-
TH-DM-DBT and 4,6-HH-DM-DBT decreased with weight
time. These results suggested that more active sites were
accessible in Pt/MNZ-5 because of the existence of meso-
pores, although Pt/MNZ-5 had a similar metal dispersion as
Pt/NZ-5 (Table 2). The finding that Pt/g-Al O had lower
Figure 1. Conversion in the HDS of 4,6-DM-DBT over Pt catalysts on
various supports.
conversion represents the yield of all desulfurized hydro-
carbon compounds. HDS conversion reached 63% for Pt/
MNZ-5 at high weight time, 31% for Pt/NZ-5, and 30% for
Pt/g-Al O (Table 2). The removal of sulfur by Pt/MNZ-5 was
2
3
2
3
twice that of Pt/NZ-5 and Pt/g-Al O .
HDS activity than Pt/MNZ-5, even though it possessed
mesoporosity and a relatively high metal dispersion, revealed
that the acidity was advantageous for the improvement of the
HDS activity.
2
3
Table 2: Metal dispersion, estimated particle size, and performance of
various catalysts.
Figure 2 shows the HDS conversion of Pd catalysts on
various supports in the HDS of 4,6-DM-DBT. Pd/MNZ-5
exhibited much higher catalytic activity than Pd/NZ-5 and Pd/
g-Al O : HDS conversion with Pd/MNZ-5 was 86% at high
[
a]
Catalyst
Dispersion [%] Particle size [nm] HDS conv. [%]
Pt/g-Al O3
74
37
35
73
27
25
37
24
23
1.3
2.6
2.7
1.3
3.5
3.8
2.5
3.9
4.0
30
31
63
21
3
2
Pt/NZ-5
Pt/MNZ-5
2
3
weight time, but only 3% for Pd/NZ-5 and 21% for Pd/g-
Pd/g-Al O3
2
Al O3 (Table 2). The sulfur removal by Pd/MNZ-5 was
2
Pd/NZ-5
Pd/MNZ-5
Pt-Pd/g-Al O3
Pt-Pd/NZ-5
2
9 times higher than for Pd/NZ-5 and 4 times higher than
86
[
b]
for Pd/g-Al O . These results indicated that MNZ-5 with
46
3
2
3
2
[
b]
mesopores was a better catalyst support in the HDS of 4,6-
DM-DBT than was NZ-5 without mesopores, thus proving
that mesoporosity in the supports played a positive role in
sulfur removal. Pd/MNZ-5 even showed a much higher HDS
activity than Pd/g-Al O , although it had a much lower metal
[
b]
Pt-Pd/MNZ-5
73
À1
À1
[
a] At weight time=9 gminmol . [b] At weight time=2.7 gminmol .
2
3
In the case of Pt/MNZ-5, trimethylcyclopentenylbenzene
TMC) and the 3,3’-dimethylcyclohexylbenzene (3,3’-DM-
CHB) and 3,3’-dimethylbicyclohexyl (3,3’-DM-BCH) isomers
were found in the final products together with the sulfur-
dispersion (Table 2), further revealing that acidity was
beneficial for enhancing the HDS activity. Therefore, the
combination of mesoporosity and acidity promotes the
efficiency of sulfur removal.
(
Angew. Chem. Int. Ed. 2008, 47, 8478 –8481
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 8479

Communications
The isomerization occurred when Pt was combined with an
acidic support. However, when Pd or Pt-Pd was used, no
isomerization was observed, probably because of their strong
hydrogenation ability. To demonstrate the contribution of
acidity on the formation of TMC isomers, 500 Pa piperidine
was added to destroy the acidity during the HDS of 4.6-
DMDBT over Pt/MNZ-5. As a result, TMC isomers were not
observed, thereby proving that acidity is the main factor for
the formation of the TMC isomers. Therefore, in the proposed
reaction scheme, 3,3’-DM-CHEB was considered as an
intermediate to form 3,3’-DM-CHB or TMC isomers.
In conclusion, the catalytic performance of noble-metal
catalysts supported on mesoporous zeolites in the HDS of 4,6-
DM-DBT was investigated for the first time. Sulfur removal
by these catalysts was much more efficient than those on
Figure 2. Conversion in the HDS of 4,6-DM-DBT over Pd catalysts on
various supports.
microporous zeolites and g-Al O . A new reaction scheme for
2
3
the HDS of 4,6-DM-DBT was proposed on the basis of the
obtained results. Further improvement in the catalytic
performance can be expected by tuning the acidity and pore
size of the mesoporous zeolites. This study proves that
mesoporous zeolites are ideal candidates of supports for
deep HDS.
To further demonstrate the efficiency of sulfur removal
for noble-metal catalysts supported on mesoporous zeolites,
the catalytic performance of the bimetallic Pt-Pd/MNZ-5
catalyst was studied. The results showed that the bimetallic
catalysts were much more active than the monometallic
catalysts, thus proving that a chemical synergism existed
between the noble metals present and that alloying had taken Experimental Section
place. HDS conversion with Pt-Pd/MNZ-5 reached 73% at
low weight time (2.7 gminmol ), but only 3% for Pt-Pd/NZ-
The mesoporous Na-ZSM-5 (MNZ-5) and conventional Na-ZSM-5
[
9i,j]
À1
(NZ-5) zeolites were prepared according to published procedures.
g-Al O was obtained from Condea. The noble-metal catalysts were
2
3
5
and 46% for Pt-Pd/g-Al O (Table 2). The sulfur removal
2 3
prepared by incipient wetness impregnation of the calcined support
with an appropriate aqueous solution of [Pt(NH ) ](NO ) (Aldrich,
was twice that of Pt-Pd/g-Al O and 24 times higher than that
2
3
3
4
3 2
for Pt-Pd/NZ-5.
9
9%) and/or [Pd(NH ) ](NO ) (Alfa, 5 wt% in Pd). After impreg-
3 4 3 2
On the basis of the obtained results, we propose a new
reaction scheme for the HDS of 4,6-DM-DBTon noble-metal
catalysts (Scheme 1). 3,3’-DM-CHEB was introduced in the
scheme, because of the observation of TMC isomers, which
should be produced by the isomerization of 3,3’-DM-CHEB.
nation, the catalysts were dried in air at ambient temperature for 3 h,
then in an oven at 393 K for 4 h, and finally calcined at 773 K for 4 h
À1
(
at a heating rate of 1 Kmin ). According to the above procedures,
monometallic 0.5 wt% Pt, 0.5 wt% Pd, and bimetallic 0.25 wt% Pt/
.25 wt% Pd were prepared. Further experimental details are given
0
in the Supporting Information.
Received: May 30, 2008
Revised: August 22, 2008
Published online: September 29, 2008
Keywords: heterogeneous catalysis · mesoporous materials ·
.
noble metal · sulfur · zeolites
[
1] R. T. Yang, A. J. Hernꢀndez-Maldonado, F. H. Yang, Science
003, 301, 79.
2
[
2] a) M. Houalla, N. K. Nag, A. V. Sapre, D. H. Broderick, B. C.
Gates, AIChE J. 1978, 24, 1015; b) J. Girgis, B. C. Gates, Ind.
Eng. Chem. Res. 1991, 30, 2021; c) D. D. Whitehurst, T. Isoda, I.
Mochida, Adv. Catal. 1998, 42, 345; d) T. Tabe, A. Ishihara, W.
Qian, Hydrodesulfurization and Hydrodenitrogenation, Wiley-
VCH, Weinheim, 1999.
[
3] a) S. D. Lin, C. Song, Catal. Today 1996, 31, 93; b) C. Song, A. D.
Schmitz, Energy Fuels 1997, 11, 656; c) L. J. Simon, J. G.
van Ommen, A. Jentys, J. A. Lercher, J. Catal. 2001, 203, 434;
d) A. Niquille-Rꢁthlisberger, R. Prins, J. Catal. 2006, 242, 207;
e) Y. Yoshimura, M. Toba, T. Matsui, M. Harada, Y. Ichihashi,
K. K. Bando, H. Yasuda, H. Ishihara, Y. Morita, T. Kameoka,
Appl. Catal. A 2007, 322, 152; f) T. Tang, C. Yin, L. Wang, Y. Ji,
F.-S. Xiao, J. Catal. 2007, 249, 111.
Scheme 1. Reaction pathway of the HDS of 4,6-DM-DBT on noble-
metal catalysts.
8480 www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 8478 –8481

Angewandte
Chemie
[
4] a) H. R. Reinhoudt, R. Troost, S. van Schalkwijk, A. D. van
Langeveld, S. T. Sie, H. Schulz, D. Chadwick, J. Cambra, V. H. J.
de Beer, J. A. R. van Veen, J. L. G. Fierro, J. A. Moulijn, Stud.
Surf. Sci. Catal. 1997, 106, 237; b) W. Robinson, J. A. R.
van Veen, V. H. J. de Beer, R. A. van Santen, Fuel Process.
Technol. 1999, 61, 103; c) H. R. Reinhoudt, R. Troost, A. D.
van Langeveld, J. A. R. van Veen, S. T. Sie, J. A. Moulijn, Stud.
Surf. Sci. Catal. 1999, 127, 251; d) F. Bataille, J. L. Lemberton, G.
Perot, P. Leyrit, T. Cseri, N. Marchal, S. Kasztelan, Appl. Catal. A
men, A. Jentys, J. A. Lercher, J. Catal. 2001, 201, 60; c) L. J.
Simon, J. G. van Ommen, A. Jentys, J. A. Lercher, J. Catal. 2001,
203, 434.
[9] a) C. J. H. Jacobsen, C. Madsen, J. Houzvicka, I. Schmidt, A.
Carlsson, J. Am. Chem. Soc. 2000, 122, 7116; b) S. Kim, J. Shah,
T. Pinnavaia, Chem. Mater. 2003, 15, 1664; c) Y. Tao, H. Kanoh,
K. Kaneko, J. Am. Chem. Soc. 2003, 125, 6044; d) Z. Yang, Y.
Xia, R. Mokaya, Adv. Mater. 2004, 16, 727; e) W. Li, A. Lu, R.
Palkovits, W. Schmidt, B. Spliethoff, F. Schꢂth, J. Am. Chem. Soc.
2005, 127, 12595; f) Y. Tao, H. Kanoh, L. Abrmas, K. Kaneko,
Chem. Rev. 2006, 106, 896; g) Y. Fang, H. Hu, J. Am. Chem. Soc.
2006, 128, 10636; h) F.-S. Xiao, L. Wang, C. Yin, K. Lin, Y. Di, R.
Xu, D. S. Su, R. Schlꢁgl, T. Yokoi, T. Tatsumi, Angew. Chem.
2006, 118, 3162; Angew. Chem. Int. Ed. 2006, 45, 3090; i) M.
Choi, H. Cho, R. Srivastava, C. Venkatesan, D. Choi, R. Ryoo,
Nat. Mater. 2006, 5, 718; j) R. Srivastava, M. Choi, R. Ryoo,
Chem. Commun. 2006, 4489; k) H. Wang, T. J. Pinnavaia,
Angew. Chem. 2006, 118, 7765; Angew. Chem. Int. Ed. 2006,
45, 7603; l) Y. Sun, R. Prins, Appl. Catal. A 2008, 336, 11.
[10] G. M. Kramer, G. B. McVicker, Acc. Chem. Res. 1986, 19, 78.
2001, 220, 191; e) T. Matsui, M. Harada, M. Toba, Y. Yoshimura,
Appl. Catal. A 2005, 293, 137; f) A. Niquille-Rꢁthlisberger, R.
Prins, Catal. Today 2007, 123, 198.
[
5] a) F. Figueras, B. Mencier, L. De Mourgues, C. Naccache, Y.
Trambouze, J. Catal. 1970, 19, 315; b) P. Gallezot, Catal. Rev. Sci.
Eng. 1979, 20, 121; c) W. M. H. Sachtler, A. Y. Stakheev, Catal.
Today 1992, 12, 283.
6] B. H. Cooper, B. B. L. Donnis, Appl. Catal. A 1996, 137, 203.
7] a) S. D. Lin, M. A. Vannice, J. Catal. 1993, 143, 539; b) S. D. Lin,
M. A. Vannice, J. Catal. 1993, 143, 563; c) S. D. Lin, M. A.
Vannice, J. Catal. 1993, 143, 554.
[
[
[
8] a) L. J. Simon, J. G. van Ommen, A. Jentys, J. A. Lercher, J.
Phys. Chem. B 2000, 104, 11644; b) L. J. Simon, J. G. van Om-
Angew. Chem. Int. Ed. 2008, 47, 8478 –8481
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
8481